If I understand you correctly, you are using a microwave beam to heat Silicon Carbide. You say the temperature distribution is not uniform. Well microwave ovens have this problem and solve it by rotating the food or putting a paddle stirrer in the microwave beam to scramble the beam. Can you deflect the microwave beam using some sort of movable vane to scatter (randomise) the beam, giving a more uniform power distribution?

I don’t understand why the power required to heat the bed is ever increasing. What is changing? Is it the Silicon Carbide decomposing for example?

Hello logbook

Do you think the scattering the beam will help in uniformly heating the bed? May be that ll help just uniformly heating the top layer not deep in the bed. The reactor is around 250 mm in diameter and 500 mm deep.SiC is a very stable material.I dont think its decomposing.
Regards
Fahims

Fahims,
understand that I don’t know exactly what your setup is, but a fixed frequency microwave beam will bounce off metal surfaces and create a standing wave pattern. There will therefore be high and low field intensities at various fixed locations within any structure. By randomising the beam you turn a fixed standing wave pattern into a multiplicity of “fixed” standing wave patterns. A node (null point) in one pattern will not be in the same position as a node in another pattern, hopefully.

Let me put some guestimates here, having no experience in this field whatsoever, based purely on “academic” thoughts. If we feed a microwave beam into a lossy material with a reflector at the end we get some sort of lossy standing wave. This is a nasty solution because as the beam travels it gets weaker. We therefore cannot assign a standing wave ratio to the field. If we could assign a standing wave ratio to the pattern, then the lossy material wouldn’t be very lossy!

If the reflector had a reflection coefficient of unity, with relatively low loss in the material, the standing wave ratio would tend to infinity. Anti-nodes having infinitely more power dissipated per unit volume than the nodes would then exist. A reflection coefficient at the end of 0.5 would give a standing wave ratio of 3 meaning three times the power in some areas compared to others. Unfortunately it would also mean that 75% of the power would be dissipated at the reflector!

This end-fed solution is really hopeless unless you get a variable frequency magnetron (if that is even possible for the huge tuning range required to move the nodes significantly).

Why don’t you sketch out how the microwave beam is distributed with respect to the SiC bed and we can think about it some more. Draw it on paper, scan it in, upload it to your own webspace (everyone gets there own free webspace from ISPs these days), and post a hyperlink.

carbide tends to be a very lossy material. what other materials have you considered?